Both Type 1 and Type 2 diabetes (T1D and T2D) result from inadequate numbers of normally functioning beta cells. Small molecule drugs that inhibit the kinase, DYRK1A, such as harmine and others, are reproducibly able to induce adult human beta cells to replicate, but at low rates (~2%/day). More recently, we have shown that adding any small molecule DYRK1A inhibitor to either a TGF-beta superfamily inhibitor (TGF?I's) or to a GLP1 receptor agonist such as GLP1 or exendin-4 markedly enhances this replicative induction to rates averaging 5-8%/day. This has been documented not only by ?markers? of replication such as Ki67 and BrdU, both in vitro and in vivo, but also by enhancing actual numbers of beta cells. Unfortunately, these drugs may not exclusively affect the beta cell, but instead may have ?off target? effects as well. Thus, diabetes researchers have effective regenerative drugs to deliver to the human beta cell, but may require a ?targeting molecule? to bring them to the beta cell. Accordingly, over the past few years, diabetes researchers have identified two classes of ?prototype beta cell targeting molecules?: a monoclonal antibody raised against the beta cell surface molecule called ?ENTPD3?; and, the GLP1-receptor class of molecules. While these molecules may or may not be perfect for beta cell targeting, they are unquestionably ?prototype? targeting molecules with which to work while the field attempts to identify more perfect beta cell targeting molecules. We have also synthesized numerous DYRK1A inhibitors and TGFbeta inhibitors and cleavable chemical linkers that enable their conjugation to any potential beta cell targeting molecule. Thus, the Aims of this application are: 1. Synthesis of TGF-beta Inhibitors With Chemical Linkers, to Complement our Novel DYRK1A inhibitor Linker Compounds, For Conjugation To Prototype Targeted Delivery Vehicles. 2. Conjugation of Harmine-Linker and TGF-beta-Inhibitor-Linker Compounds to Two Prototype Targeting Molecules: GLP1 Receptor Agonists and ENTPD3 Monoclonal Antibodies. 3. Definition of Long Term Efficacy, Specificity and Safety of the Harmine-Linker and TGF-beta Inhibitor- Linker Conjugates in vivo in Human Islet Engraftment Models. These goals are both achievable and directly responsive to the goals of aims of the NIDDK. If GLP1 receptor agonists and/or ENTPD3 MAbs prove suboptimal for beta cell targeting, the approaches and molecules developed here can readily be extended and adapted to any future more specific human beta cell targeting molecule.